View Usage

Usage metrics show how your models are used — token consumption, API requests, RPM/TPM, and cost trends. Use them to monitor daily activity, investigate spikes, and compare usage across models.

​

Home overview

The Home page is a quick usage snapshot for the selected time range. Summary cards show total tokens, API requests, average RPM, and average TPM; the chart below switches between token usage, total cost, and API requests, and can be filtered by model. Use it for routine checks — whether traffic is normal, which models are active, and how usage compares with the previous period.

​

Metrics dashboard

Open Metrics for deeper analysis: per-model breakdowns of token usage and API requests, plus RPM and TPM trends over time. Use preset ranges for recent activity, or a custom date range to review incidents, monthly usage, or cost changes.

​

Investigate usage

When usage looks unexpected, filter by model and time range to isolate the traffic, then cross-check it against your API keys — pause suspicious keys, lower their quotas or rate limits, and create a new key if needed.